How the Race was Lost: Boost Mode Beta Testing

May 5, 2020

LAST UPDATED May 20, 2020

Zwift’s latest update includes a new FutureWorks event type called “Boost Mode”. It’s already been the source of no small amount of controversy, with some Zwifters complaining they don’t want “fake” watts/artificially-boosted Strava segment times, Zwift should spend their time developing features Zwifters are requesting, and “we don’t want Mario Kart – we want a realistic cycling simulator.”

Fair enough. But I would push back on those complaints:

I don’t want fake watts: this isn’t like riding an ebike! Your own watts charge your “battery”, then you release those watts when desired. Zwift has been clear in saying there are “no free watts” in Boost Mode. In-game segment PRs are disabled in Boost Mode events. As for Strava segments – my guess is the boosts aren’t long enough to affect segment times, except on short sprints.
Zwift should spend time on other features: Zwift employs around 300 people, and there are always many projects underway in various stages. Just because Zwift’s developers are working on one thing doesn’t mean they’re neglecting another. Boost Mode is a great example of this – I first heard about it back in December 2019, when Zwift was talking about releasing it! So it was in development for months prior to this, along with a slew of other features.
We don’t want Mario Kart: here’s the thing: some people do! Zwift is different things to different people, which is why events have options. Boost Mode isn’t going to be used in every race – far from it. But it may be used in some races, and you can take part in those, or not.

Yesterday I saw that the Jarvis Club has a private test event set up: “Boost Mode Test 1a”. 10 Laps of Crit City’s Downtown Dolphin, and the first-ever live test of Boost Mode. I had to sign up! I wasn’t entirely sure how Boost Mode worked, but I figured it would be simple enough. Turns out I was both right… and wrong. Boost Mode takes Zwift strategy to the next level! Here’s the story of today’s Boost Mode race.

The Warmup

Today’s warmup was a bit longer than my usual, because the race wasn’t until 8am and I was antsy to ride. So I got on around 7am after chewing three pieces of caffeine gum and slapping some PR lotion on the legs. I had to spend a little extra time tweaking my trainer setup, because I was testing out some new Garmin Vector 3 pedals and had my bike hooked to Monica’s rocker plate and Tacx Neo trainer.

A few hard efforts got my heart rate up to 160, then it was time to head to the Crit City start pens. Let’s race!

Welcome to Boost Mode

When I clicked to join the event, I was greeted with a “Welcome to Boost Mode!” screen. This introduced me to Boost Mode, and let me set my options:

Here’s how Boost Mode works:

You have three different “Boost Types” to choose from, and you have to choose one. You only get to pick one of these for your event, and you can change it until the event begins. Think of these as on-bike batteries which you charge yourself.
For this test event, the three Boost Types were:
- Efficient: 50w boost for up to 90 seconds. Requires 50w to charge. Adds 11lb to your weight throughout the event.
- Balanced: 200w boost for up to 17 seconds. Requires 100w to charge. Adds 16.5lb to your weight throughout the event. This is the boost I selected.
- Power Burst: 500w boost for 5 seconds. Requires 150w to charge. Adds 22lb to your weight throughout the event.
Each boost type has a different number of “stages”. You can store up 4 boosts with “Efficient” mode, 2 boosts with “Balanced”, and only one boost in “Power Burst”. Tapping once on the charge or discharge button will charge/discharge one stage. Tapping multiple times will charge/discharge as many times as you tap.
During the event, buttons in Companion and on-screen let you decide when you want to charge your Boost, and when you want to use it.
Small green icons show next to rider names when they are using a boost. You have no way of knowing which boost type they selected, or when they are charging their boost.

The Start

The clock hit zero and our small band of ~15 B’s headed out of the pens. And that’s when I encountered my first problem: the A’s showed up! I hadn’t looked closely at the event details, and didn’t realize we’d be mixing in with the other categories. (Have I mentioned how much I dislike racing with other categories?)

I pushed extra hard to hang with the front group, knowing the longer I could hang with the A’s, the better my chance at a race win. But I was off the back almost immediately – not surprising considering the A’s included heavy hitters like Kim Little, Casey Schumm, Nathan Guerra, and James Hodges!

My First Charge

Once the failed fight for the front finished, I turned my attention to Boost Mode. I needed to charge my first boost! So as I began descending the twisties, I tapped the battery icon in Companion. This reduced my power by 100 watts while it charged, as indicated by the “-100w” graphic below the power number at the top-left of the screen:

Zwift doesn’t change your power numbers shown on screen (pure watts or w/kg number) when you’re charging or using a boost. And they shouldn’t! We all want to see accurate power.

Crucially, Zwift also doesn’t change the power numbers it records while you’re charging or boosting. So your final fit file will be accurate, showing the actual power you put out during the event.

Timing Bug

There was one bug that Zwift will need to fix, though. I didn’t notice this until after the race, when I was watching my recording… but it took the same amount of time (8.5 seconds) for me to charge my boost as it took for the boost to discharge. Why does this matter? Because charging only cost me 100w, but discharging boosted me by double that (200w)! So I was getting 100 free watts for 8.5 seconds every time I used a boost!

I chatted with Wes Salmon at HQ about this – and he made it very clear that their goal here is for boost mode to give no free watts. Which is exactly how it should be. So this is just a bug which will be fixed in short order – probably by the end of today, according to Wes.

Wes also said they’re building in a bit of a “charging tax”, so the watts you get when activating your boost should be a little less than what you put in – like charging a real battery.

The Middle

As the race continued, I started paying more attention to what was going on around me, both in terms of the overall race situation, and how other riders were using Boost Mode. Our B group was quite scattered, with a few riders up the road with some A’s, and me with a small pack of B’s in what I assumed was the second group. (This is one reason why I don’t like racing with mixed categories – it makes it harder to track who is where!)

I had my eye on Chris Pritchard, who was up the road with some of the A’s. I figured he (as well as others) were live-streaming this effort, and you know how that goes: when you know the world is watching, you push yourself to the max!

My boost charging and discharging fell into a more regular pattern as the race continued. I would charge on the descending twisties, because descents are where it’s easiest to keep speed high even if watts drop off. Conversely, I would use my boost on the brick prime climb, because climbs are where wattage boosts help your speed the most. This had the effect of making my effort more “even” over each lap.

Extended Animation

But just because it evened out my effort doesn’t mean it made the race boring! The opposite was true, in fact: Boost Mode really animated this race, in ways I have never experienced before on Zwift. Having a boost in my back pocket gave me the gumption to attack hard on the climb or bridge up to riders ahead, when I would have otherwise sat in with my group. Instead of the typical Crit City race of attrition, I found myself in a constantly-changing small group of riders as we caught some up the road, dropped others, and got caught by others using boosts to chase us down!

The 200w boost was definitely noticeable – it made a huge difference in terms of effort up the prime climb, and really boosted my speed on the flats. The -100w hit when charging was also noticeable, but not too bad when done on the descent.

There was an entirely new level of strategy at play here as riders tried to leapfrog groups or attack hard and drop others. All of the Boost Mode madness was mixed in with the standard powerup fun, too, so it was possible to (for example) use a feather plus a 200w boost on the prime climb.

The Finish

In the final lap I found myself with a pack of two other B riders: Mike Ellis and Chris Pritchard. There were just a couple B’s well up the road, as far as I could tell… so we were fighting for a podium spot or two.

Mike had been putting in some hard attacks throughout the race, and I figured Chris had a strong sprint. I had held onto an aero helmet powerup for the final sprint, and charged both of my boosts. So I had 17s of a 200w boost to use, plus a 15-second powerup. What was my best strategy in the final meters of this race? Now you can see how Boost Mode brings another level of thought to Zwift racing! And my poor addled race-brain was having a hard time deciding the best move.

Given how I’d seen Mike attacking thus far, I figured he would go hard on the prime climb and try to stay away. Pritchard, on the other hand, had been racing smart – sitting in the draft. So I decided I would try to follow the wheel of anyone who attacked hard on the climb, using one Boost if needed. And I would save my other Boost plus aero powerup for the final sprint.

Mike jumped on the prime climb as expected – and he jumped hard! Watching the replay, he must have used a Boost, because he flew away quickly. I was gapped, even with my extra Boost power. Pritchard used a Boost as well as a feather powerup to make it up the climb easily, and we traded pulls as we chased Mike on our way down the descending twisties to the final sprint.

Timing the boost and powerup wasn’t easy, with Mike just within reach and Chris right on my tail. I hit both of my “assists” probably earlier than I should have, getting out ahead of Chris then passing Mike before Chris came roaring past in the final meters. I was right – he did have a strong sprint!

In the end I got 4th place. One nice thing about Jarvis Club private events – no sandbaggers! Our in-game results matched ZwiftPower’s results.

See my activity on Strava >
See my activity on Zwift >
See race results on ZwiftPower >

Watch My Recorded Race

Takeaways

My big takeaway from this event is that Boost Mode is going to be a big deal in Zwift racing. It will never be used for all races, and it shouldn’t be, because some racers want a more “pure” race experience (no powerups, either). But I believe once Zwift gets Boost Mode dialed in, Zwifters will find they really enjoy the way it animates races and adds strategic options.

There’s plenty of strategy and variability in outdoor bike racing which we don’t get on Zwift: wind, mechanicals, pack positioning, braking/surging, etc. Boost Mode, in my opinion, will bring some much-needed variability to Zwift races.

My second takeaway is that I’m happy with how Zwift is letting Jarvis Club members test these new FutureWorks features. This is exactly how it should be done – let experienced Zwifters test stuff out and provide feedback so things can be polished then released to the entire Zwift community. For Boost Mode in particular, Wes mentioned that next week they’ll have more events, and the numbers for the boosts will be tweaked, as HQ dials in what works best for charging wattage, boost wattage, and charge times.